Brigham Young University Brigham Young University

BYU ScholarsArchive BYU ScholarsArchive

Theses and Dissertations

2010-07-09

Investigation of Potential Trapping Bias in Malaise Traps Due to Investigation of Potential Trapping Bias in Malaise Traps Due to

Mesh Gauge, in Two Habitats Mesh Gauge, in Two Habitats

David Jensen Betts Brigham Young University - Provo

Follow this and additional works at: https://scholarsarchive.byu.edu/etd

Part of the Biology Commons

BYU ScholarsArchive Citation BYU ScholarsArchive Citation Betts, David Jensen, "Investigation of Potential Trapping Bias in Malaise Traps Due to Mesh Gauge, in Two Habitats" (2010). Theses and Dissertations. 2561. https://scholarsarchive.byu.edu/etd/2561

This Thesis is brought to you for free and open access by BYU ScholarsArchive. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of BYU ScholarsArchive. For more information, please contact scholarsarchive@byu.edu, ellen_amatangelo@byu.edu.

Investigation of Potential Trapping Bias of Insects in Malaise Traps

Due to Mesh Gauge, in Two Habitats

David J. Betts

A thesis submitted to the faculty of Brigham Young University

in partial fulfillment of the requirements for the degree of

Master of Science

Nikki L. Hanegan, Chair, C. Riley Nelson Russell B. Rader

Department of Biology

Brigham Young University

August 2010

Copyright © 2010 David J. Betts

All Rights Reserved

ABSTRACT

Investigation of Potential Trapping Bias of Insects in Malaise Traps

Due to Mesh Gauge, in Two Habitats

David J. Betts

Department of Biology

Master of Science

Malaise traps are a common tool for collecting insects used by many researchers.

Although there have been variations in the models and materials used for Malaise traps, the potential for sampling bias due to mesh gauge has been explored inadequately. This study compared coarse and fine mesh Townes model Malaise traps in two habitats on the Grand Staircase-Escalante National Monument. The two habitats next to the Lick Wash trailhead were defined by dominant vegetation type – sagebrush and grasses or Piñon-Juniper. We collected from three sites per habitat type, over three consecutive days in June in both 2006 and 2007. A pair of Malaise traps consisting of one coarse mesh and one fine mesh trap was used at each site in order to compare differences in the diversity and in the average size of individuals collected by each type of Malaise trap. We measured diversity using both presence-absence data such as richness scores and Jaccard’s Index of Similarity, and abundance-based measures of comparison, including Simpson’s Index of Diversity and non-metric multidimensional scaling. We identified all individuals according to Order, and because of our interest in flies and their abundance, we further identified the Diptera to the Family level. Average insect size was determined by categorizing individuals according to one of 14 distinct size-classes. In sum, 71 samples totaling approximately 62,500 insects were identified and sized. Because we sampled from two adjacent habitats, we also discuss beta diversity across the sample sites. Although mesh-size appears to have a significant effect on the diversity of the catch according to some tests, not all of our analysis agrees. In addition, the gain in the amount of diversity collected by incorporating both mesh-sizes may not be worth the costs of that kind of sampling. Other means of collection may adequately make up that difference. Habitat on the other hand was a clear marker for difference in diversity. Size was not found to be significant overall, but there still may be reasons to examine the effect of mesh-size with respect to the Hymenoptera.

Keywords: Malaise traps, sampling bias, insects, Diptera (flies), Grand Staircase-Escalante National Monument, Utah, body-size, richness, abundance, similarity indices, Simpson’s Index of Diversity, Jaccard’s Index, beta diversity

ACKNOWLEDGMENTS

My thanks first go to the members of my committee, Drs. Nikki L. Hanegan, C. Riley

Nelson, Russell B. Rader. Thank you for your knowledge in discussing the ideas of this study

and for the material support you provided in collecting and analyzing this data. Thanks also to

Utah’s Biodiversity Experiences for Students and Teachers (UBEST) Program, which was

funded by the Grand Staircase-Escalante National Monument, the Utah State Office of Education

and Brigham Young University, and its many participants who provided the opportunity and

funding to collect this data. The Charles Redd Center for Western studies also provided some

funding to help sort the collected samples.

Thanks also to Dr. Dennis L. Eggett of the Department of Statistics at Brigham Young

University for performing the statistical analysis with the SAS software.

A combination of many students provided hundreds of hours in the lab sizing and sorting

the thousands of insects that were collected for this study, including but probably not limited to:

Andrew Wilmore, Trevor Hess, Marta Dagvasuren, Sara (Morrison) Wursten, Karen Froerer,

Laura (Price) Chisholm, Roger Stimpson, Brad Mortenson, Denton Davenport, Scott Murdock,

Samuel Zmolek and of course the many UBEST participants.

Both my parents and my wife’s parents have been an unending source of support. Laura

and I will forever be in your debt for your generosity. Thanks to the rest of our families for at

least trying to understand my answer to your question of “Why bugs?”

And finally, thanks to my girls, Laura, Maggie, and Chloe, for continually giving me

something to look forward to each day and at the end of this project.

iv

Table of Contents

Table of Contents ........................................................................................................................... iv

Introduction ..................................................................................................................................... 1

Cataloging Biodiversity ............................................................................................................... 1

Mitigating Sampling Bias for More Accurate Estimates of natural populations ........................ 1

Malaise Traps: a common method of collection ......................................................................... 2

Current use ............................................................................................................................... 2

Past research with respect to sampling bias ............................................................................. 3

Research Question .......................................................................................................................... 4

Influence of Mesh-size ................................................................................................................ 4

Diversity Across Habitats ............................................................................................................ 5

Materials and methods .................................................................................................................... 5

Sampling Methods ....................................................................................................................... 5

Malaise traps ............................................................................................................................ 5

Sampling location ..................................................................................................................... 6

Habitat assessment ................................................................................................................... 6

Trapping sites ........................................................................................................................... 7

Sampling schedule .................................................................................................................... 8

Sample-bias Analysis .................................................................................................................. 8

Taxonomic diversity ................................................................................................................. 8

Insect size ................................................................................................................................. 9

Results ........................................................................................................................................... 10

Overall diversity ........................................................................................................................ 10

Richness ..................................................................................................................................... 13

Similarity comparisons .............................................................................................................. 13

v

NMDS plots............................................................................................................................ 14

SIMPER ................................................................................................................................. 17

Insect Size .................................................................................................................................. 18

Mixed Model Procedures in SAS .............................................................................................. 18

Hymenoptera ............................................................................................................................. 20

Discussion ..................................................................................................................................... 20

Differences in Diversity............................................................................................................. 20

Presence-Absence Estimates .................................................................................................. 20

Abundance-based estimates ................................................................................................... 21

Body-size ................................................................................................................................... 21

The average size of insects ..................................................................................................... 21

Differences Due to Mesh-size ................................................................................................ 22

Beta Diversity ............................................................................................................................ 22

Conclusion .................................................................................................................................... 23

Influence of Mesh-size on Diversity ......................................................................................... 23

Influence of Mesh-size on Average Body-size.......................................................................... 24

Hymenoptera May Be an Exception .......................................................................................... 24

Patterns of Beta Diversity .......................................................................................................... 25

Literature Cited ............................................................................................................................. 27

1

Introduction

Cataloging Biodiversity

Part of the work of Biology is to catalogue communities and their components (Wilson

1992; Magurran 2004). Whether your concern is with conservation and preservation, taxonomy

and systematics, ecological and community relationships, or a single taxon or subset of taxa there

remains the basic need of getting an estimate of community composition.

Insects are the most abundant taxon within the animal kingdom (Wilson 1987; Dial and

Marzluff 1988; Wilson 1988; Wilson 1992) Insects play significant roles in their communities

and with respect to humanity. There are familiar negative roles as pests to humans, crops, and

livestock, or as vectors for disease. Insects also have positive characters as pollinators,

decomposers, and as significant links in the food chain(Brown 2005). From a human standpoint,

insects have also been used as bioindicators (Lenat 1988; Burgio and Sommaggio 2007), as

models in genetics and evolution (Mitchell-Olds 2001; Celniker and Rubin 2003), and multiple

other aspects of scientific research (Wigglesworth 1985; Papaj and Lewis 1992; Finlay, Thomas

et al. 2006). Inadequate sampling due to sampling bias could very well result in a

misunderstanding of not only the diversity and biogeography of insects, but perhaps humanity

may overlook other potential explanatory patterns and models (Brown 2005; Fraser, Dytham et

al. 2008).

Mitigating Sampling Bias for More Accurate Estimates of natural populations

As scientists, we also have to recognize the limitations of time and money with respect

trying to maximize efficiency of collecting. Past research has shown that with insects, more than

one method of trapping is needed in order to collect a high percentage of the available taxa

2

(Covell and Freytag 1979; Darling and Packer 1988; Brown 1995; Brown and Feener 1995;

Faulds and Crabtree 1995; Campos, Pereira et al. 2000; Suh, Spurgeon et al. 2003; Bartholomew

and Prowell 2005; Brown 2005; Holusa and Drapela 2006; Fraser, Dytham et al. 2008; Mazon

and Bordera 2008; van Hennekeler 2008; Missa 2009). These and other studies recognize that

there are limitations to each collection method, passive or active.

Regardless of the sampling method(s) used, efforts should always be made to reduce bias

in order to obtain better estimates of diversity of natural populations. Even with highly

successful trapping methods – success being measured strictly by the number of individuals

captured – it should not be assumed that every available taxon is captured or that each taxon

captured is captured proportionally to the natural population (Magurran 2004).

If a sampling bias is present, an understanding of this bias will help us to better mitigate

these limitations and increase the economy of our sampling efforts (Darling and Packer 1988;

Campos, Pereira et al. 2000).

Malaise Traps: a common method of collection

Current use

Our concern for this paper is the Malaise trap. The first model was developed by René

Malaise, a Hymenopterist, in 1934 (Malaise 1937), and the trap was well adopted into use by at

least the 1960s . By the mid 1970s, several trap designs and modifications had been made and

experimented with and some consensus has been arrived in using the Townes model (Townes

1972; Matthews and Matthews 1983). The Townes model is the model we used in this study.

The Malaise trap is used in particular for Hymenopterans (Darling and Packer 1988;

Noyes 1989; Saaksjarvi, Haataja et al. 2004; Bartholomew and Prowell 2005; Fraser, Dytham et

3

al. 2008; Mazon and Bordera 2008) and Dipterans (Cerretti, Whitmore et al. 2004; Brown 2005;

Fritzler 2006; Marinoni, Marinoni et al. 2006; Ziegler 2006; Tschorsnig, Domingo-Quero et al.

2007; Wilson 2007; Ackland 2008). Several other orders have also been listed as taxa of interest

in studies citing the use Malaise traps, including: Lepidoptera (Covell and Freytag 1979;

Tangmitcharoen, Takaso et al. 2006; Liska and Modlinger 2007; Dapporto and Strumia 2008),

Coleoptera (Cook ; Ulyshen, Hanula et al. ; Cunningham and Murray 2007; Noguera, Chemsak

et al. 2007; Abdullah, Sina et al. 2008; Linzmeier and Ribeiro-Costa 2008), Psocoptera (Read ;

Chan and Yang 2005); Hemiptera (Quednau ; Gonzon, Bartlett et al. 2006; Inoue, Goto et al.

2007), Thysanoptera (Olsen and Midtgaard 1996), and even some apparently stubborn

Collembola (Riedel, Marinoni et al. 2008).

The Malaise trap finds preference because it is a passive form of collection with minimal

maintenance (Matthews and Matthews 1971). Catches can be collected and traps refreshed easily

after periods as long as one or two weeks. Malaise traps are also popular because of the large

numbers of individuals it collects. (Brown and Feener 1995; Bartholomew and Prowell 2005)

Past research with respect to sampling bias

There has been some recognition that not all Malaise traps are equal with regards to what

they collect. Some studies have focused on the shape or model of the trap (Platt, Caldwell et al. ;

Roberts 1972; Townes 1972; Matthews and Matthews 1983), others have asked questions with

regards to materials (Darling and Packer 1988), age of materials (Roberts 1975), color (Roberts

1970; Barbosa, Henriques et al. 2005), the addition of baits (Blume, Miller et al. ; Davis, Zwick

et al. ; Roberts ; Schreck, Kline et al. 1993; Rohrig 2008), or placement (Noyes 1989; Schreck,

Kline et al. 1993; Suh, Spurgeon et al. 2003; Grimbacher and Stork 2007; Irvine and Woods

2007; Vance, Smith et al. 2007; van Hennekeler 2008). Even with the variety of studies

4

comparing Malaise trap efficiency, the term Malaise trap is most frequently generalized in the

literature regardless of the model or materials from which it is made, as evidenced by the sources

cited in this paper and as noted by others (Campos, Pereira et al. 2000).

We have found just one study besides our own that attempts to address the possible bias

due to the gauge of the mesh used to construct the trap (Darling and Packer 1988). This study

does recognize the potential influence that mesh-size may have on the diversity of insects

collected, and notes the possibility that body-size of the insects may be a part of the mechanism

influencing that outcome. Unfortunately their experimental design was based on two traps that

still differed in color. There are also other potential limitations due to the fact that Darling and

Packer used only two individual traps.

Research Question

Influence of Mesh-size

Our research question is this: Are the kinds of insects collected in a Malaise trap different

due to mesh-size? We examine this question via two means. First, is the diversity of insects

collected different according to mesh-size? Second, are the sizes of insects different according

to mesh-size? As with the Darling and Packer paper (1988), we are going to use insect body-size

as a stand-in for potential influence of mesh-size. As to what aspect of behavior, perception or

other factor is reflected by the size of insects collected per mesh-size we cannot say with this

study.

We expected that the Malaise samples would show a difference due to habitat. We

recognize that there are potential differences due to the year in which samples were collected, but

we will only briefly discuss comparisons of year to help provide context to the two main

5

variables (mesh-size and habitat) that we are most interested in. Recognizing the possibility that

some insects may be able to pass through the larger mesh-size, our expectation of a difference if

we were to find one would favor more small individuals trapped with the fine mesh.

Diversity Across Habitats

We chose to sample in two habitats in order to provide greater strength to our

comparisons of the influence of mesh-size. With the data we collected, we will also discuss

some patterns we found with respect to beta diversity.

Materials and methods

Sampling Methods

Malaise traps

Our samples were collected with standard commercially available Townes’ model

Malaise traps, purchased from www.santetraps.com. Color, age, and dimension of all the traps

were equal. The traps did differ according to the gauge of the holes in the mesh, which are listed

as “coarse mesh” and “fine mesh” by the vendor. The holes in the coarse mesh are

approximately 1.0mm in diameter (Fig. 1) while the holes in the fine mesh are between 0.3mm

and 0.5mm (Fig. 2).

6

Fig. 1 - Coarse mesh (scale = 1mm) Fig. 2 - Fine mesh (scale = 1mm)

Sampling location

Our sampling location was in the Grand Staircase-Escalante National Monument near

Kanab, Utah which is located in Southern Utah between the Zion, Bryce Canyon, and Capitol

Reef National Parks and the Grand Canyon. We sampled from the sagebrush and grass dominant

community (Flat) along Skutumpah Road at the head of Lick Wash, and within an adjacent plant

community dominated by woody plants such as Piñon Pines, Utah Junipers, and various shrubs

on the hillside (Hill) rising above the wash. Both habitats are open to grazing.

Habitat assessment

We completed vegetative surveys within 16 m2 blocks adjacent to each of the sample

sites. These surveys were done on a presence absence basis only. We were unable to identify

all of the plants in each site. Nevertheless, comparing the number and kinds of woody species

alone we were confident in the distinction between the two habitat sites. The grasses and forbs

that were positively identified only strengthen this distinction. Overall, essentially all of the

plants were indentified to the Family level, with many identified at the species level. The

overwhelming majority of plants were identified to at least the level of Genus.

7

Trapping sites

The two habitats are separated by a graded dirt road, a distance of 500m and an elevation

difference between the two habitats ranging from 20m to 40m. The Flat sites were all within two

meters in elevation of each other, but the Hillside sites covered this range of twenty meters in

order to have adequate spacing while staying within the same habitat type based on dominant

plant composition.

Three replicate sites were selected per habitat with a 50m distance between each site and

its nearest neighbor to ensure independence of sample sites while maintaining habitat similarity.

Pairs of coarse and fine mesh Malaise traps (one trap of each mesh-size per site) were used in

each of the two habitats for a total of six trapping locations and twelve traps.

All traps were placed with a north-south orientation for the long axis of the traps out of

convenience (it is easy to find the north-south axis with even a simple compass) and for

consistency over sampling seasons. Whether the coarse or fine net was on the north or south side

of the center post was randomly selected at each site each year via coin toss. The location data

for each site was recorded at the center post between the paired traps for each site. (Table 1)

Lick Wash Sample Sites Site Latitude Longitude Elevation (m) Flat 1 37.36641 -112.18986 1903 Flat 2 37.36687 -112.18961 1904 Flat 3 37.36715 -112.19012 1905 Hill 1 37.36200 -112.18877 1925 Hill 2 37.36164 -112.18900 1934 Hill 3 37.36111 -112.18921 1945

Table 1 - GPS locations of sample sites.

8

Sampling schedule

Our samples were collected approximately every twenty-four hours in an attempt to

encompass a complete diurnal cycle of activity for three consecutive days in June each year. The

samples were collected on the 20th, 21st and 22nd in both 2006 and 2007. The traps were set up

and collected from by the participants in the Utah Biodiversity Experiences of Students and

Teachers (UBEST) program; by both the secondary science teachers enrolled in the program and

the instructors and students from Brigham Young University in charge of directing the UBEST

program.

Sample-bias Analysis

Taxonomic diversity

Our first means of comparing potential effects of mesh-size is to compare the diversity of

the samples collected. Trap samples have been sorted down to the Order level, including the

occasional non-insect arthropods. One trap sample is defined by the combination of day,

location, and mesh-size – e.g. 20 Jun 2006, Flat habitat, coarse mesh. Because of the abundance

of flies in the samples and our interest in this group, the Diptera were sorted to the Family level.

Similarity will be measured using both presence-absence data and similarities based on

abundance.

We used a simple richness score as our presence-absence measure per trap. To balance

out against richness, we use Simpson's Index of Diversity (1 − 𝐷; 𝐷 = ∑ 𝑛𝑖(𝑛𝑖𝑠𝑖 −1)𝑁(𝑁−1)

) as one of our

abundance measures. Richness by nature is strongly influenced by rare taxa, while Simpson's

Index is less easily swayed by these taxa. Richness and Simpson’s Index of Diversity were

compared using mixed model analysis with the SAS statistical software.

9

Bray-Curtis estimates are another abundance-based diversity measure used to compare

diversity between trap samples. Primer 6 was used to calculate the majority of these indices and

much of analyses of diversity. Within Primer 6 we performed non-metric multi-dimensional

scaling tests (NMDS), analysis of similarities (ANOSIM), and similarity percentages (SIMPER)

to test for the similarity between mesh-size (Coarse vs. Fine), habitats (Flat vs. Hill), and years

(2006 vs. 2007) in addition to trap site locations. We compared trap sites simply as a part of our

discussion on beta diversity.

In addition, we will consider pair-wise comparisons of similarity using Jaccard's Index of

similarity in order to examine the relationship between the flora and fauna of our sample sites, as

well as to discuss the nature of presence-absence versus abundance measures when discussing

diversity.

Insect size

To compare the potential for size-bias, individuals were sized using a template with

circles of various diameters. (Fig. 3) The smallest circle has a diameter of 1mm and the largest

circle has a diameter of 60mm. All individuals were placed into one of the 14 categories. Each

individual was categorized according to the smallest diameter circle within which their entire

body, including wings and legs, would fit. No individuals exceeded the largest circle. Because

the size categories were neither in strict linear increments or followed a true logarithmic scale,

our size estimates are based on the mid-point between categories.

10

Fig. 3 – Sizing sheet used for categorizing individuals (not to scale).

Average insect size per trap was compared in the same mixed model tests as richness and

Simpson’s Index of Diversity, using the SAS software. Because the traps were set up in pairs,

coarse and flat nets are not truly independent. The mixed models analysis performed modified

paired t-tests using the differences between the pairs of coarse and fine mesh traps, blocking the

samples by day (six categories), and maintaining the variable for habitat (two categories).

Results

Overall diversity

71 of the possible 72 samples were sorted according to our protocol. One sample was

lost before sorting was completed. A total of 62,497 individuals were identified and sized. The

average number of individuals per trap was 880 insects, with a range from 239 individuals to

2,382 individuals. There are 13 orders, two of which were non-insect orders (Table 2 & Fig. 4).

11

As Brown (2005) noted with his citations, Malaise traps collect mostly flies. The Diptera

clearly formed the overwhelming majority of the individuals collected (Fig. 4), with the Family

Chloropidae alone constituting more than half of all individuals collected (Fig. 5). Even with

such a high percentage of the total catch, chloropids have a range of 19 individuals to 2021

individuals per trap.

Arachnida

Araneae 22

Ixodida 53

Insecta

Coleoptera 316

Collembola 1

Diptera 50471

Hemiptera 744

Homoptera 2128

Hymenoptera 3687

Lepidoptera 4771

Neuroptera 177

Orthoptera 13

Psocoptera 8

Thysanoptera 106

Table 2 - Number of individuals per Order

Fig. 4 - Relative abundance per Order

Diptera81%

Lepidoptera 8%

Hymenoptera6%

Homoptera 3%

Hemiptera 1%

Remaining taxa1%

12

The total richness for the entire study included 55 taxa, once Diptera had been sorted to

the Family level. The seven most abundant taxa have over 92 percent of the total number of

individuals collected (Fig. 5). The remaining 48 taxa are individually less than 1% of the total

number of individuals, and 23 of those are individually less than 0.1% of the total number of

individuals collected. There were seven singletons represented by a single individual: one

collembolan and six Diptera families. All other taxa were collected from two or more samples,

with the seven most abundant taxa collected in every trap.

Fig. 5 - Relative abundance of Orders and Families

Chloropidae 56%

Cecidomyiidae 12%

Lepidoptera 8%

Dolichopodidae 6%

Hymenoptera6%

Homoptera 3%

Hemiptera 1%

Tachinidae 1%

Anthomyiidae 1%

Remaining 46 Taxa

6%

13

Richness

Richness

Arthropods Plants

Overall Richness 55 44

Mesh-size Coarse 49 --

Fine 52 -- Habitat

Flat 50 18 Hill 52 38

Year 2006 50 --

2007 50 -- Table 3 - Richness for arthropod and plant taxa

Plants Site Richness Flat 1 16 Flat 2 18 Flat 3 15 Hill 1 38 Hill 2 28 Hill 3 20

Table 4 - Plant taxa richness per site

Similarity comparisons

Jaccard's Indices of Similarity Flat Vs. Hill (Arthropods) 0.85

Flat Vs. Hill (Plants) 0.18

Coarse Vs. Fine 0.84 2006 Vs. 2007 0.82

Table 5 – Pair-wise similarity comparisons using Jaccard's Index of Similarity

14

NMDS plots

Fig. 6 - Similarity comparison of mesh-size (Coarse vs. Fine)

15

Fig. 7 - Similarity comparison of Year (2006 vs. 2007)

16

Fig. 8 - Similarity comparison of Habitat (Flat vs. Hill)

17

Fig. 9 - Similarity comparison of trap sites

ANOSIM p-values

Factor p-value Mesh-size 0.029 Year 0.004 Habitat 0.001 Trap Site 0.001

Table 6 - Corresponding p-values to the NMDS plots

SIMPER

SIMPER lists the taxa most responsible for the characterization of the trap samples, and we

compared results across each of the four factors represented in the NMDS plots (Figs. 6, 7, 8 &

9). The results of all comparisons largely followed the relative abundance of all 55 taxa (Fig. 5).

Chloropidae was nearly always the taxa responsible for the largest percentage of similarity. In

18

order to reach the 90% level of characterization using SIMPER, there was only moderate

variation away from the overall abundance levels.

Insect Size

0

5000

10000

15000

20000

25000

0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 12.5

17.5 25 45

Fig. 10 - Insect abundance per size category. (Size in mm).

Mixed Model Procedures in SAS

Although all 71 samples were used with the analysis with the analysis using Primer 6, the

sample without a corresponding pair (Flat 1, Fine, 21 Jun 2007) was left out of the analysis using

SAS.

19

Richness

Type 3 Tests of Fixed Effects

Num Den

Effect DF DF

F Value Pr > F

Habitat 1 10 0.53 0.4814

Least Squares Means

Standard Effect Habitat Estimate Error DF t Value Pr > |t|

Habitat Flat 0.4118 0.8942 10 0.46 0.655

Habitat Hill -0.5 0.8691 10 -0.58 0.5778

Table 7 - Analysis of the effect of mesh-size on richness

Simpson's

Type 3 Tests of Fixed Effects

Num Den

Effect DF DF F Value Pr > F

Habitat 1 10 0.18 0.6783

Least Squares Means

Standard Effect Habitat Estimate Error DF t Value Pr > |t|

Habitat Flat 0.009488 0.04462 10 0.21 0.8359

Habitat Hill -0.01709 0.04336 10 -0.39 0.7018

Table 8 - Analysis of the effect of mesh-size on Simpson's Index of Diversity

Average Size

Type 3 Tests of Fixed Effects

Num Den

Effect DF DF F Value Pr > F

Habitat 1 10 2 0.1876

Least Squares Means

Standard Effect Habitat Estimate Error DF t Value Pr > |t|

Habitat Flat 0.3922 0.3386 10 1.16 0.2736

Habitat Hill 1.066 0.3351 10 3.18 0.0098

Table 9 - Analysis of the effect of mesh-size on Average body-size per trap

20

Hymenoptera

Total individuals (Hymenoptera) per mesh type

Coarse 757 Fine 2119

Individuals (Hymenoptera) per trap

Standard Standard

95.0% LCL

95.0% UCL

Variable Count Mean Deviation Error of Mean of Mean Individuals Coarse 35 21.62857 10.20405 1.7248 18.12336 25.13379 Individuals Fine 36 58.86111 29.71787 4.952978 48.80603 68.91619

Table 10 - Number of individual Hymenoptera per trap Average Size (mm) per trap - Hymenoptera

Standard Standard

95.0% LCL

95.0% UCL

Variable Count Mean Deviation Error of Mean of Mean AvgSize_Coarse 35 6.791719 2.308826 0.390263 5.99861 7.584829 AvgSize_Fine 36 4.051429 1.392027 0.232004 3.580435 4.522423

Table 11 - Average size (mm) of Hymenoptera per trap

Discussion

Differences in Diversity

Presence-Absence Estimates

The Jaccard’s Index of Similarity of 0.18 for the two plant communities (Table 5) and the

difference in richness scores across habitats (Table 3) support our initial visual assessment of

distinct habitat types. In comparison, the corresponding indices for the arthropod communities

are much more similar.

Arthropod richness is essentially equal both within and between the three variables we

listed: mesh-size, habitat, and year (Table 3). It is no surprise that the mixed model analysis

gave no evidence of a difference in richness due to mesh-size; p-value = 0.48 (Table 7). The

Jaccard’s Indices for these same variables also show a high degree of similarity, with all three

21

indices between 0.82 and 0.85 (Table 5). From a presence-absence standpoint, the arthropod

communities across each variable are highly similar.

Abundance-based estimates

When we examine abundance on an per trap basis, the mixed model analysis of

Simpson’s Index of Similarity gives no evidence that there is a difference due to mesh-size; p-

value = 0.68 (Table 8). With both the Simpson’s Index and with richness, these scores are

calculated with no reference to the other samples or groups of samples. Two samples could have

equal richness scores or Simpson’s Indices, while being comprised of entirely different taxa. In

order to compare abundance-based diversity relative to other samples we used the tools in

Primer 6.

The Stress value of 0.09 for the NMDS plots (Figs. 6, 7, 8 & 9) gives evidence that much

of the diversity of the insects we collected is well represented by the two-dimensional plots. The

ANOSIM p-values of all the variables displayed with the NMDS plots are significant (Table 6).

Including abundance levels that have relationships across samples has modified our view of what

is similar. There is a statistically discernable difference due to these variables.

Body-size

The average size of insects

On a purely individual basis, insects are small (Fig. 10). Overall more taxa were at the

smaller end of the size scale on average and the most abundant taxa were prominently

categorized in these lower size-classes.

One size-class pattern that may not be easily apparent due to the scale of the graph is a

second concentration of individuals around the size class of 12.5mm. The most abundant Orders

22

besides Diptera (Lepidoptera and Hymenoptera) (Fig. 4) had many individuals in these upper

size-classes. In addition, several fly families, but especially the calypterate flies which had some

of the more abundant taxa (Tachinidae, Anthomyiidae, Muscidae, etc.) (Fig. 5) also had size

distributions that more closely centered near the 12.5mm size class. While most of the insects

we sampled are small, there appears to be a second mode of size distribution close to 12.5mm.

Differences Due to Mesh-size

The mixed model analysis in SAS was our only means of comparing the effect of mesh-

size on the diversity of body-size. Although the individual t-tests give some support that body-

size is influenced by mesh-size, overall we do not have enough evidence (p-value = 0.19) to

conclude that mesh-size affects the average size of insect collected per trap (Table 9).

Beta Diversity

Both the data comparing plant communities (Tables 3, 4 & 5) and the data comparing

arthropod communities (Figs. 8 & 9) show how easily beta diversity can change over a short

distance. As noted earlier, our habitats were 500m apart and immediately adjacent.

While our intent was to sample from two distinct habitats, the NMDS plot based on

sample site location (Fig. 9) provides strong evidence that the Hill habitat was not homogenous.

The groupings of the samples from the Flat sites in this figure are indistinguishable, while the

proximity of the sites in the Hill habitat to the Flat habitat is reflected by the groupings in the

NMDS plot. Differences in elevation correlate with this same plot. The lowest sample sites

from an effectively equal elevation group together. The remaining samples sites fall in line in

the plot according to increasing elevation.

23

The data with respect to the plant community also reflects this pattern of proximity to the

Flat habitat. The richness scores of each sample site should be noted (Table 4). The plant

richness scores from the Flat sites show a strong sense of equality, while the highest richness

score of the Hill sites is found in the site nearest the Flat habitat and the remaining scores

decrease with the corresponding increase in distance from the Flat habitat. We suggest that some

of this heterogeneity of the Hill habitat is due to an intermediate zone between the plants at the

lowest elevation of our study and the highest elevation. For example, the presence of sagebrush

(Artemesia tridentata) decreased according to elevation. Site Hill 1 was placed within the first

grouping of Piñons and Junipers where much sagebrush was still present, while site Hill 3 was

entirely devoid of sagebrush and Hill 2 was at some intermediate level. We do not have specific

abundance data for the plants, but perhaps the high richness score of site Hill 1 is due to the

overlapping of two plant communities.

Conclusion

Influence of Mesh-size on Diversity

The main purpose of this study was to determine the influence of mesh-size on sampling

bias, because of the potential influence on experimental design. We found evidence that mesh-

size influences the diversity of the overall sample of insects collected, although not all forms of

our analysis agreed (Tables 3 & 4 as compared to Figs. 6, 7 & 8). The corresponding ANOSIM

p-value to the NMDS plot comparing mesh-size (Fig. 6) gives evidence that the insect diversity

collected by the two mesh types are not the same. The large overlap with respect to diversity as

present in the NMDS plot is perhaps evidence enough that although the differences in diversity

are statistically significant, these differences may not be biologically significant. It should be

24

recognized that our sampling periods were limited to three days each summer and that we

sampled a total of six sites. Accounting for the differences we found in future sampling efforts

may be too labor intensive for the small gains that would be achieved with a detailed

experimental design with respect to mesh-size. The same may be said with respect to the

differences due to year. The differences we found in the diversity of our samples might be

overcome at a lower cost with more traps, the combination of other trapping methods, an

extended trapping season, or any combination of those or other variables.

Influence of Mesh-size on Average Body-size

We found no evidence to support the hypothesis that mesh-size influences the average

body-size of insects collected overall. The results from the mixed model analysis give evidence

that perhaps we might find a true difference if we could increase the power of our test (Table 9),

but again that kind of study is most likely not worth the effort to find a difference that might not

be biologically significant. The estimated average differences in body-size were 1.1mm for the

Hill habitat and 0.39mm for the Flat habitat. For insects as a whole, the meaning of those

differences still needs to be explored.

Hymenoptera May Be an Exception

There is one caveat to these assessments of the influence of mesh-size – the

Hymenoptera. Hymenoptera are the one Order of insects that is probably of interest to those

who use Malaise traps as often as or more often than the Diptera. Our analysis was with regard

to overall diversity and average body-size across all taxa collected; Diptera being the largest

portion. Knowing that Hymenopterists would be the next most likely group interested in this

study, we compared the sampling patterns of the Hymenoptera alone.

25

The data collected do not meet the requirements of a t-test, but the descriptive statistics

we include in Tables 10 and 11 do provide some evidence that further examination of our data

may be warranted. First of all, the fine mesh Malaise traps collected a total of 2119 individuals

(excluding the Formicidae); while the coarse mesh Malaise traps collected only 757 individuals

(excluding the Formicidae). The means and their confidence intervals for the number of

individuals captured per trap (Table 10), and the average body-size per trap (Table 11) do not

overlap. The fine mesh Malaise traps not only collected more individuals of all categories, but

collected insects belonging to higher size categories that were not collected by the coarse mesh

Malaise traps. These results agree with the results from Darling and Packer that the fine mesh

may be more effective than the coarse mesh at collecting the smaller Hymenoptera (1988).

Again, this data has not been properly analyzed, but we think that there is sufficient

evidence for those interested in this taxon to take a closer look. Likewise, for those interested in

collecting small taxa, e.g. parasitic wasps, or for those who are trying to reduce their work load

by excluding smaller taxa, there may be undiscovered valid reasons to consider the mesh-size of

the Malaise traps you use.

Patterns of Beta Diversity

As biologists, we want to find what is different. Differences within the context of

overarching relationships and similarities are the keys to discovery of pattern and process. To

continue the advancement of our biological understanding, being able to predict where

differences are found would improve the way we study Biology. We could better diversify and

standardize our sampling. We would more closely understand where to look for the next

unknown species, population, community, etc.

26

In this study we see a change in diversity at some level across all variables. With only

500m between sites, we do not have to go far to find different plant communities. In an even

shorter distance within the Hill habitat, we can see further evidence of changes in beta diversity

when we use the abundance data for the arthropods. Even our limited scope in time of three days

per year displays how differences can be found (Fig. 7).

In addition, this same design and locality was used in the summer of 2005, but that data

was not included in this analysis because several samples were lost in the intervening time

between collection and identification. In 2005 sampling was done during the first week, rather

than the third week of June. With preliminary identification, community composition already

appeared to be different from the 2006 and 2007 data. In 2006 and 2007, we did not catch a

single individual from the family Tipulidae. With our incomplete sorting of the 2005 data, we

have found nearly ten tipulids per sample. Time also appears to provide a broad opportunity for

collecting new levels of diversity.

As researchers we should recognize that when concerned with plants or insects, we may

not have to go far geographically in order to find different communities. We should also

consider that to discover the composition of a community, our sampling seasons should match

the natural community cycles. Our brief sampling does not define the boundaries of these

seasons, but does support how much can change over a short period of time.

27

Literature Cited

Abdullah, F., I. Sina, et al. (2008). "The ground beetle fauna (Coleoptera: Carabidae) of Kenyir water catchment, Terengganu, Peninsular Malaysia." Pakistan Journal of Biological Sciences

Ackland, D. M. (2008). "Diptera Stelviana. A dipterological perspective on a changing Alpine landscape: results from a survey of the biodiversity of Diptera (Insecta) in the Stilfserjoch National Park (Italy). Volume 1: 4.3.04 Anthomyiidae."

11(21): 2478-2483.

Studia Dipterologica Supplement

Barbosa, M. G. V., A. L. Henriques, et al. (2005). "Species diversity and similarity between sites in Tabanidae (Insecta: Diptera) of a terra firme forest (Adolpho Ducke Reserve) in Central Amazonia, Brazil."

16: 107-116.

AmazonianaBartholomew, C. S. and D. Prowell (2005). "Pan compared to malaise trapping for bees

(Hymenoptera: Apoidea) in a longleaf pine savanna."

18(3-4): 251-266.

Journal of the Kansas Entomological Society

Blume, R. R., J. A. Miller, et al. "Trapping tabanids with modified malaise traps baited with CO2."

78(4): 390-392.

Mosquito NewsBrown, B. (1995).

(1): 90. Phorid flies, malaise traps, and pan traps: Collections in a tropical biodiversity

surveyBrown, B. V. (2005). "Malaise trap catches and the crisis in Neotropical dipterology."

. American

EntomologistBrown, B. V. and D. H. J. Feener (1995). "Efficiency of Two Mass Sampling Methods for

Sampling Phorid Flies (Diptera: Phoridae) in a Tropical Biodiversity Survey."

51(3): 180-183.

Contributions in ScienceBurgio, G. and D. Sommaggio (2007). "Syrphids as landscape bioindicators in Italian

agroecosystems."

(459): 1-10.

Agriculture, Ecosystems & EnvironmentCampos, W. G., D. B. S. Pereira, et al. (2000). "Comparison of the efficiency of flight-

interception trap models for sampling Hymenoptera and other insects. ."

120(2-4): 416-422.

An. Soc. Entomol. Bras.

Celniker, S. E. and G. M. Rubin (2003). "The Drosophila melanogaster genome." 29(3): 381-389.

Annual Review of Genomics and Human Genetics

Cerretti, P., D. Whitmore, et al. (2004). "Survey on the spatio-temporal distribution of tachinid flies - using Malaise traps (Diptera, Tachinidae)." 229-256.

4: 89-117.

Chan, M. L. and J. T. Yang (2005). Psocoptera Survey by Using Malaise Trap in Central Taiwan

Cook, S. P. "Beetle (Coleoptera) diversity in mixed pine-hardwood stands in the Ouachita highlands five years following tree harvests."

.

Southwestern entomologistCovell, C. V., Jr. and P. H. Freytag (1979). "The malaise trap as a means of sampling butterfly

populations in Kentucky."

27(3/4): 269.

Journal of the Lepidopterists' SocietyCunningham, S. A. and W. Murray (2007). "Average body length of arboreal and aerial beetle

(Coleoptera) assemblages from remnant and plantation Eucalyptus forests in southwestern Australia."

33(3): 153-161.

OecologiaDapporto, L. and F. Strumia (2008). "A rapid assessment of Lepidoptera biogeography in the

Tuscan Archipelago using economical Malaise traps."

151(2): 303.

Italian Journal of Zoology (Modena) 75(2): 219-224.

28

Darling, D. C. and L. Packer (1988). "Effectiveness of Malaise traps in collecting Hymenoptera: the influence of trap design, mesh size, and location." Canadian entomologist

Davis, H. G., R. W. Zwick, et al. "Perimeter traps baited with synthetic lures for suppression of yellowjackets in fruit orchards."

120(8/9): 787-796.

Environmental entomologyDial, K. P. and J. M. Marzluff (1988). "Are the Smallest Organisms the Most Diverse?"

2(4): 569. Ecology

Faulds, W. and R. Crabtree (1995). "A system for using a Malaise trap in the forest canopy."

69(5): 1620-1624.

New Zealand EntomologistFinlay, B. J., J. A. Thomas, et al. (2006). "Self-similar patterns of nature: insect diversity at local

to global scales."

18: 97-99.

Proceedings of the Royal Society B. Biological sciencesFraser, S. E. M., C. Dytham, et al. (2008). "The effectiveness and optimal use of Malaise traps

for monitoring parasitoid wasps."

273: 1935.

Insect Conservation and DiversityFritzler, C. J. (2006). "Diversity of Syrphidae (Diptera) in central Appalachian

forests."

1: 22-31.

Proceedings of the Entomological Society of WashingtonGonzon, A. T., C. R. Bartlett, et al. (2006). "Planthopper (Hemiptera: Fulgoroidea) Diversity in

the Great Smokey Mountains National Park."

108(4): 964-976.

Transactions of the American Etnomological Society

Grimbacher, P. S. and N. E. Stork (2007). "Vertical stratification of feeding guilds and body size in beetle assemblages from an Australian tropical rainforest."

132(3/4): 243-260.

Austral Ecology

Holusa, J. and K. Drapela (2006). "Yellow sticky boards: a possible way of monitoring little spruce sawfly (Pristiphora abietina) (Hymenoptera: Tenthredinidae)."

32(1): 77-85.

Journal of Forest Science

Inoue, T., H. Goto, et al. (2007). "Cicadas (Hemiptera: Cicadidae) collected with Malaise traps set in natural forests and plantations in the northeastern part of Ibaraki Prefecture."

52(1): 13-21.

Shinrin Sogo Kenkyujo Kenkyu HokokuIrvine, K. M. and S. A. Woods (2007). "Evaluating shading bias in Malaise and window-pane

traps."

6(4): 249-252.

J. Acad. Entomol. Soc.Lenat, D. R. (1988). "Water Quality Assessment of Streams Using a Qualitative Collection

Method for Benthic Macroinvertebrates."

3: 38-48.

Journal of the North American Benthological Society

Linzmeier, A. M. and C. S. Ribeiro-Costa (2008). "Seasonality and temporal structuration of Alticini community (Coleoptera, Chrysomelidae, Galerucinae) in the Araucaria Forest of Parana, Brazil."

7(3): 222-233.

Revista Brasileira de Entomologia [Rev. Bras. Entomol.]. Vol. 52

Liska, J. and R. Modlinger (2007). "Investigation of lepidopteran fauna (Insecta, Lepidoptera) of montane spruce forest in the massive of Trojmezna Mt. in the Bohemian forest."

52(2): 289-295.

Silva Gabreta

Magurran, A. E. (2004). 13(1): 57-64.

Measuring Biological DiversityMalaise, R. (1937). "A new insect trap."

. Malden, MA, Blackwell Publishing. Entomologisk Tidskrift

Marinoni, L., R. C. Marinoni, et al. (2006). "Most abundant species of Syrphidae (Diptera) collected during two years with Malaise traps in Parani, Brazil."

58: 148-160.

Revista Brasileira de Zoologia

Matthews, R. W. and J. R. Matthews (1971). "The Malaise trap: its utility and potential for sampling insect populations."

23(4): 1071-1077.

Michigan Entomologist 4(4): 117-122.

29

Matthews, R. W. and J. R. Matthews (1983). "Malaise traps: the Townes Model catches more insects." Contributions of the American Entomological Institute

Mazon, M. and S. Bordera (2008). "Effectiveness of two sampling methods used for collecting Ichneumonidae (Hymenoptera) in the Cabaneros National Park (Spain)."

: 428.

European Journal of Entomology

Missa, O. (2009). "Monitoring arthropods in a tropical landscape: relative effects of sampling methods and habitat types on trap catches [electronic resource]."

105(5): 879-888.

Journal of insect conservation

Mitchell-Olds, T. (2001). "Arabidopsis thaliana and its wild relatives: a model system for ecology and evolution."

13(1): 103-118.

Trends in Ecology & EvolutionNoguera, F. A., J. A. Chemsak, et al. (2007). "A faunal study of Cerambycidae (Coleoptera)

from one region with tropical dry forest in Mexico: San Buenaventura, Jalisco."

16(12): 693-700.

Pan-Pacific Entomologist

Noyes, J. S. (1989). "A study of five methods of sampling Hymenoptera (Insecta) in a tropical rainforest, with special reference to the Parasitica."

83(4): 296-314.

Journal of Natural History

Olsen, A. J. and F. Midtgaard (1996). "Malaise trap collections of thrips from the islands Haoya and Ostoya in Oslofjorden, South Norway."

23(2): 285 - 298.

Norwegian J. Entomol.Papaj, D. R. and A. C. Lewis (1992).

43: 63-68. Insect Learning: Ecological and Evolutionary Perspectives

Platt, J., J. S. Caldwell, et al. "An easily replicated, inexpensive malaise-type trap design."

, Springer.

Journal of entomological scienceQuednau, F. W. "A survey of aphid fauna (Homoptera: Aphididae) in Quebec hardwood forest

stands by means of Malaise trap catches."

34(1): 154.

Rapport d'information LAU-X - Laurentian Forest Research Centre

Read, R. W. J. "Psocoptera activity as monitored by a Malaise trap." (67).

Entomologists' monthly magazine

Riedel, R., R. C. Marinoni, et al. (2008). "Spatio-temporal trends of insect communities in southern Brazil."

114(pub. 1979): 94.

Journal of EntomologyRoberts, R. H. "Tabanidae collected in a Malaise trap baited with CO2."

5(6): 369-380. Mosquito News

Roberts, R. H. (1970). "Color of Malaise trap and the collection of Tabanidae." (1): 52.

Mosquito News

Roberts, R. H. (1972). "The effectiveness of several types of Malaise traps for the collection of Tabanidae and Culicidae."

30(4): 567-571.

Mosquito NewsRoberts, R. H. (1975). "Influence of trap screen age on collections of tabanids in -Malaise

traps."

(4): 542.

Mosquito newsRohrig, E. (2008). "Comparison of Parasitic Hymenoptera Captured in Malaise Traps Baited

with two Flowering Plants, Lobularia maritima (Brassicales: Brassicaceae) and Spermacoce verticillata (Gentianales: Rubiaceae) [electronic resource]."

(4): 538.

Florida entomologist

Saaksjarvi, I. E., S. Haataja, et al. (2004). "High local species richness of parasitic wasps (Hymenoptera: Ichneumonidae: Pimplinae and Rhyssinae) from the lowland rainforests of Peruvian Amazonia."

91(4): 621-627.

Ecological entomologySchreck, C. E., D. L. Kline, et al. (1993). "Field evaluations in malaise and canopy traps of

selected targets as attractants for tabanid species (Diptera: Tabanidae)."

29: 735.

Journal Of The American Mosquito Control Association 9(2): 182.

30

Suh, C. P. C., D. W. Spurgeon, et al. (2003). "Evaluation of sticky traps for monitoring cotton fleahopper movement into cotton." Proceedings Beltwide Cotton Conferences

Tangmitcharoen, S., T. Takaso, et al. (2006). "Insect biodiversity in flowering teak (Tectona grandis L.f.) canopies: comparison of wild and plantation stands."

2003: 1444-1447.

Forest Ecology and Management

Townes, H. (1972). "A light-weight Malaise trap." 222(1-3): 99-107.

Entomological NewsTschorsnig, H. P., T. Domingo-Quero, et al. (2007). "Tachinidae (Diptera) from the Caldera de

Taburiente National Park, La Palma (Canary Islands)."

83(9): 239-247.

GraellsiaUlyshen, M. D., J. L. Hanula, et al. "Using Malaise traps to sample ground beetles (Coleoptera:

Carabidae)."

63(1): 43-51.

Canadian entomologistvan Hennekeler, K. (2008). "A comparison of trapping methods for Tabanidae (Diptera) in North

Queensland, Australia."

137: 251.

Medical And Veterinary EntomologyVance, C. C., S. M. Smith, et al. (2007). "Differences Between Forest Type and Vertical Strata in

the Diversity and Composition of Hymenpteran Families and Mymarid Genera in Northeastern Temperate Forests."

22(1): 26-31.

Environmental Entomology [Environ. Entomol.]. Vol. 36

Wigglesworth, V. B. (1985). 36(5): 1073-1083.

Insect PhysiologyWilson, E. O. (1987). "The Little Things That Run the World (The Importance and Conservation

of Invertebrates)."

, Springer.

Conservation BiologyWilson, E. O. (1988). The Current State of Biodiversity.

1(4): 344-346. Biodiversity

Wilson, E. O. (1992).

. E. O. Wilson and F. M. Peter. Washington, National Academy Press. vol. 1: 3-18.

The Diversity of LifeWilson, R. D. (2007). "Altitudinally restricted communities of Schizophoran flies in

Queensland's Wet Tropics: vulnerability to climate change [electronic resource]."

. New York, W.W. Norton & Co.

Biodiversity and conservationZiegler, J. (2006). "The "Diptera stelviana" project. A dipterological perspective on a changing

alpine landscape."

16(11): 3163-3177.

Studia Dipterologica

13(2): 195-202.